My interest in collaboration as an avenue to dialogue--a term which I use here to designate the negotiation of socially constructed meaning by the interaction of multiple voices--was instigated by an exposure to the theory of Mikhail Bakhtin. Although Bakhtin's primary focus was on the novel, his discussions of discourse in the novel, which he identified as either monologic or dialogic, fit nicely within the framework of the composition theories of social constructivism and social-epistemic rhetoric (q. v.Clark, 1990, Phelps, 1988). Additionally, Bakhtin's conflation of written and spoken texts, the "utterance," provides a possible theorization of the nature of hypertext. Bakhtin's approach provides an avenue for Catherine F. Smith's (1994) suggestion in "Hypertextual Thinking" that "[t]alk, theorized as conversation and analyzed as discourse, may provide the models of interaction that we need, in order to improve the design of hypertext systems and to extend the reach of its applications" (p. 281).
I see collaboration as a method of showing students that meaning is socially constructed and that writing is a "form of activity inseparable from the wider social relations between writers and readers [that is] largely unintelligible outside the social purposes and conditions in which [it is] embedded" (Eagleton, 1983, p. 216); Eagleton's definition is, not coincidentally, very similar to Bakhtin's view of dialogue as an action which occurs between self and other--as Michael Holquist describes it, dialogue "expresses the general condition of each speaker's addressivity, the situation of not only being preceded by a language system that is 'always already there,' but preceded as well by all of existence" (1990, p. 61). Because knowledge is socially constructed through the dialogue of language, it is therefore advisable to engage in dialogue with mulitple viewpoints in order to protect against the stagnating effect of homogenous thought (or language); Bakhtin makes this clear in his treatment of heteroglossia.